Guys, was there ever a resolution to this topic? Traxion, did you ever get rid of that smell? I have the same problem on mine.

If so, what values did you use on the O2 constants?

 

Trax, this may seem like a basic newbie question, but it has roots in fluid flow...
This topic has touched on fresh air entrainment into the exhaust at idle. Very cool stuff. This is possibly a very basic...or a very complex...question. How do a set of ported and polished heads (polished exhaust runners) factor into this equation? Larger x-sectional area for fluid flow = better head flow = more fresh air in exhaust??? What about the lack of a turbulent layer at the fluid boundary between the runner surface and the flow channel? how would this effect the whole pie so to speak?

 

nope......just more questions.

 

It depends upon the low lift flow of the head and charge motion. During overlap the valves are barely off the seat, and the intake port has yet to start flowing toward the chamber. The exhaust is either being scavaged or it can be stagnent (this depends upon the exhaust configuration).

During low lift any intake flow travels along the top of the chamber. Which also puts it right in line with the closing exhaust valve.

RBob.

 

So you could theoretically end up with a small amount of air/fuel in the exhaust flow path. Notice I said THOERETICALLY. If the engine was getting too much fuel, this could cause things like blown up mufflers and trashed O2 sensors...right?

 

You are always going to have raw fuel in the exhaust, especially
with an non effecient cam. The smell can be reduced but not really rid of.
Excessive raw fuel can ruin a muffler, deffinetly ruin a cat. But you really get a problem when you have a bad exhaust leak allowing fresh o2 in to burn that fuel in your exhaust,

This condition is also affected by stroke length and even rod length.
The longer the stroke and the longer the rod allow more dwell time at TDC, Wasting alot of your effects of scavaging right out the exhaust as opposed to in the cylinder.. Hence your 02 reads
all that fresh 02 skewing what it thinks is correct.

 

This is not really true. You guys have assumed the times when the offset for A.I.R. system putting air in the manifold is being applied.

Go back and look at the code. Near as I can tell, the 100 mV offset is used nearly all the time, except when in divert mode. If it is in cold mode, or warm mode, the A.I.R. offset is removed from the table vaues. If the air is diverted, it is not subtracted.

My personal opinion is that this is an error. I think it should only be used when it is putting air in the manifold. I couple of simple code changes could fix this.

Scot

 

That's what I thought. Makes alot more sense........
(530-580)-100(for a.i.r)= 430-480mv. That falls in line with the hardware.

guestions are fun

 

I dunno, but the code looks OK to me. Divert active being where the A.I.R. pump output is 'diverted' to atmosphere. Maybe GM has it wrong, 1992 Helm manual, p. 6E3-C6-2 (VIN F & VIN 8), 1st column:

"Divert Mode. - Both soleniods are de-energized which opens the convertor valve, allowing air to take the path of least resistance, i.e., out the divert/relief tube to the atmosphere."

The manual then goes on to detail when this may happen.

RBob.

 

Hmmm...might have to get busy and add a couple of LEDs to the ECM bench, to indicate AIR solenoid operation.

Never thought I'd need to do that, seeing as how nothing in my driveway actually has an AIR pump.

 

It may defy logic, in it being used for keeping the converter at temp., or getting it to temp., as quickly as possible.

It would be interesting to see what a WB mounted pre converter would indicate.

 

*raw fuel*?, ya well, maybe some, but enough to really matter?.

Actually it's the ratio of rod lenght to stroke that makes a difference in the dwell. For a given stroke, increasing the rod lenght, increases dwell. While a long rod motor can have advantages in the leverage applied to the crank, it also demands better breathing ability.

 

To Make a HP engine smell differently, you bet!

Thats what I said
A long rod engine is hurt by less scavaging effect resulting in a slight loss in HP but gain some Torque due to the dwell @ TDC.
Thus making it pull more raw air and fuel into the exhaust..

But a long rod engine also has advantages not so much on crank pressure but piston sidewalls, Sometime harder on a crank because of the heavier rod. But a long rod engine alot of times uses a lighter piston because of wrist pin location

 

I typed this from the Chevrolet Service Manual into the header of the A.I.R. subroutine in my disassembly:

code:

*
*
***************************************************************************
* A.I.R. solenoids
*
* From the GM manual, page 6E3-C6-2:
*
* Air from the A.I.R. pump enters the body of the control valve and
* builds pressure against the port valve. The port solenoid controls
* the port valve which performs a switching operation by directing
* airflow from the the A.I.R. pump to the exhaust ports (solenoid
* energized) or the converter valve (solenoid de-energized). The
* converter valve which directs air to the catalytic converter (solenoid
* energized) or diverts air to the atmosphere (solenoid de-energized).
*
* Air enters the body of the valve from the pump. Air pressure builds
* against the control valve and for:
*
* * Cold mode - The port solenoid is energized which in turn opens
* the port valve and allows flow to the exhaust ports.
*
* * Warm mode - The port solenoid is de-energized and the converter
* solenoid energized which closes the port valve and
* keeps the converter valve seated, thus forcing flow
* past the converter valve and to the converter.
*
* * Divert mode - Both solenoids are de-energized which opens the
* converter valve, allowing air to take the path of
* least resistance, i.e., out the divert/relief tube
* to the atmosphere.
*
* Air is diverted to the atmosphere under the following conditions:
*
* * Rich operation
* * When the ECM recognizes a problem and sets a trouble code
* * During deceleration
* * During high RPM operation when air pressure is greater than the
* setting for the internal relief valve
*


Notice that the port (first) solenoid controls whether the air goes to the manifold or to the second solenoid switch. So if the first solenoid is active, air goes to the manifold. If non-active, it goes to the second switch, and if the second switch is active, air then goes to the CAT. If both are non-active it bypasses the manifold, and goes to the second switch, where it then gets diverted to the atmosphere.

The assumption is that the solenoid that controls the port (first) solenoid is the one that lets the code know whether to subtract the offset from the O2 test limits. This makes sense, as the air is added in front of the O2 sensor when it is in the manifold. So if that is the way it is, when the port is active, the code skips the subtract. If it is non-active (air to the second switch), it subtracts the offset. This means that the codes uses O2 table values as-is, if air is going to the manifold, and subtracts it, if the air is going to the CAT or divert.

Now, if the divert solenoid controls the decision to subtract the offset or not, this is non-sensical, as the air in either case is added after the O2 sensor, where it has no affect on the O2 sensor reading.

Now comes the screwy part in the code. The decision in the code to adjust the O2 limits is controlled by the divert solenoid status. So my thinking is that they use the wrong bit to select whether to subtract the offset or not. They should use the port solenoid status. Then all of the table values make sense, just like 11sorbust pointed out.

Scot

 

So, as I see it, you just confirmed what I originally said (without even adressing my initial comment about IDLE O2 constants). First of all, at idle the ECM should be using the idle O2mV constants ... not the tables. For AUJP that means 570-600mV at idle. It is only at cruise (actually idle constant limits) that the ECM will use the table values. The table values are more in-line with what you would expect GM to use since they are closer to the stoichiometric O2 value (which is what 11sOrBust was suggesting .... that 450-500mV should be used by the ECM). However, I don't see that being true since there is a 600mV constant for Idle.

The way you explained the operation of the AIR solenoids is exactly as I understand it. The new piece of information that you have brought to the table is that you think that the code should be subtracting the AIR R/L Difference when air is being injected into the manifolds (as I originally pointed out) but that it is NOT doing this ... because GM made a mistake and used the wrong bit. So, that means that for AUJP that the ECM really is using 600mV at idle (unless air is being pumped to the cat).

Either way, I still hold on to what I originally said. The code uses 600mV for idle. It may subtract from that (as I also pointed out on several occassions - most notably the AIR thread I made), but it is using ~600mV at idle.

See the attachment early in this thread which is from the hac showing idle constants and cruise table.

Tim

 

I dug up some old data and viewed it in Datamaster ... which allows you to look at the status bits. I can watch each status bit of the AIRMW. Assuming you are counting bits as 0 through 7 (and not 1 through 8) ... it is bits #1 and #2 that represent the 2 solenoids for AIR management.

According to the anht_hac AND according to Datamaster ...
Bit #1 is Air Dvtd
Bit #2 is Air Port

Most of the time, bit #1 is active (de-energized) except for WOT and Deceleration. I find it hard to believe that GM would be diverting air to the cat that much ... it has to be going to the manifolds. Furthermore, it is when bit #2 is set that Air is going to the cats ...

... and this DOES happen on startup according to my datamaster logs. Given that the code actually checks for bit #2 ...

Tim

 

are you 100% sure about that??

 

I would think that someone with air physically off car, but still enabled in code, using a wideband and a scan tool, could settle the big mystery.

 

No, I am not 100% sure. Scot said it wasn't getting subtracted. I started my examination of the code in an attempt to prove him wrong. GM makes a judgement call based on bit 2. Bit 2 appears to be for when air is going to the cats. See the code that I posted above.

EDIT - I'm changing some thoughts. Let's look at the code again ...

It depends on what BRSET does. If BRANCH if SET means that branching only occurs when b2 is set to 1 then that means that 100mV is always getting subtracted (except for when AIR is going to the Cat). That makes a lot of sense and would explain everything! Anybody care to elaborate on BSET? That makes a lot of sense because b1 is almost always active ... thus, 100mV would always be getting subtracted. However, 100mV would not get subtracted when air is being injected at the cats ... which would cause a richer mixture that would then burn MORE in the cats with the extra air (getting the cats hotter quicker!).

Tim

 

from a 68hc11 manual

 

Thanks, but I have access to that manual and have already reviewed the command. I just don't understand what they are trying to say.

Example:

BRSET L003F,#2,LC9D0

I know that this means Branch to LC9D0 if something (#2) is set at location L003F.

Does this ultimately mean that a branch occurs if L003F is ...

0000 0100
OR...
0000 0000

I am thinking it is 0000 0100? If this is true then Scot is right and GM did make a mistake.

Tim

 

Trax, in the instruction BRSET L003F,#2,LC9D0, the #2 is a bit mask, not the bit number. It is: 0000 0010, being a decimal 2 in binary, is bit 1. The '#' before the 2 defines it as an immediate value.

So the instruction will BRA to LC9D0 when bit 1 is set in RAM location L003F: xxxx xx1x, the x's being 'do not care'.


All, as far as the operation of the A.I.R. system and the subtraction of the O2 offset, we are assuming that we know what GM intentions are. Looking at Scot's post as to when air is pumped to port (manifold), it is only done during cold engine operation. Is the ECM even in closed loop at that time?

If not, then it doesn't matter what the offset is. If it is in closed loop then subtracting out the offset may be a good thing. This seems to be agreed upon by all posters.

Once in warm mode the air is pumped to the cat-con and the O2 offset is subtracted out. This makes sense if one wishes not to melt the cat-con. It also explains why some of the O2 constants are as high as they are.

There are times I've looked at some of the things GM has done and wondered why. Sometimes it takes thinking about the EPA & CAFE regulations to help make sense of it.

Look at the times air is in divert and the O2 offset is not subtracted out. Hmm . . .

RBob.

 

Such as? All I can tell you right now is from idle to off-idle there is a massive swing in the AFR. Anything off idle seems to be sitting right around 14.7 (averaging the data at that point) with BLM's close enough not to worry about. I'll have to disable the air in the code again to see what changes.

As for the LM1 and the .45 setting they use, they are just making the output duplicate that of the stock o2, centering at stoich at .45 It could be that you have the wire hooked up to the wrong output and thats causing the problem. I'm not having any issues with the .45 crossing point, its running pretty much the same as with the stock o2.

 

About fuel smell...

I read something a while back concerning the oxygenate additives they are using. MTBE has gone to the back shelf (Finally!), and most places they are using a different additive. Out here in Cali they have gone to (I think 10%) ethanol from MTBE. Hopefully it'll do less damage to the UST's and pollute the groundwater less... [rant]only 10 years to fix that problem[/rant] Anyway, I wouldnt be the least bit surprised if the oxygen supplying additive they are using is some sort of organic that stinks when burned. IIRC, doesnt the cornholio fuel stink? Thought thats what I read.

 

rbob, took the words out of my mouth.

a quick/easy way to see this is to look at LCA67 (page 169 of PDF HAC):

LCA67: BRSET L003F,#$80,LCA79 ; BR IF b7, DECEL ENLEAN ACTIVE

'course, there is no 80th bit, the mask is actually:

1xxx xxxx

also, I did notice a bit of an inconsistancy in the hac which might be confusing:

LCA67: BRSET L003F,#$80,LCA79 ; BR IF b7, DECEL ENLEAN ACTIVE

LCAC6: BRSET L003F,#2,LCACD ; BR IF AIR NOT DIVERTED, 0 = DIVERT

notice that the mask is specified differently, even though the meaning is the same. the full BIN instruction for each is:

12 3F 80 0E

12 3F 02 03
(opcode, operand, mask, relative offset (where to branch to))

same opcode, same way of specifying the mask. Probably just a typo in the HAC but it confused me until I looked at a .BIN to see it was the same thing (i thought there might have been different ways to specify the mask, such as directly by bit (#2) or a regular mask (#$02) but I wanted to be sure before I posted.) I thought maybe trax was picking up the bit mask differently because of the #2 instead of #$2.

 

Ok, guys, you've lost me on the way here.... I really need to learn more about how it works, but I'm learning

But I need to change the Idle O2 values when I removed the AIR to get the same "desired" O2 voltage as before at Idle?

 

Ignore this thread for what you want to do. Read the following thread...

https://www.thirdgen.org/techbb2/sho...hreadid=214321

Tim

 

I have removed the AIR the way you described, and after been reading this post agin I understand.

But I maybe going to have trouble going through emission test with my new engine, just like to be prepared what to do with the romulator if it wont pass......and changning the 600mV to lower could help a little, right?

/N.

 

so whats going on here...any progress? RBob?

 

A little more info about AIR:

 

In a earlier post
"The ecm isn't looking to maintain 14.7, it's looking to average it.
The converter NEEDS periods of rich and lean to work. So they have things set so the ecm has to go so far rich and so far lean, hence the low voltage and high voltage O2 values.
If they were looking to maintain exactly 14.7 the the would use a real tight range based on the Stoich voltage where the sensor toggles from reading rich to lean, ie if 14.7 was .45, they might use, .44 and .46, to tighten the range of variance."

Does this then mean if you are not running a cat you can tighten up the O2 values?

 

Yes, can set the O2 window wherever the engine runs the best. Can also run open loop.

RBob.

 

Old starting post and good info

 

Years later,it's been along time since anyone has come back to this topic

From this entire thread, and trying to digest it, (IF I learned anything from it) for those of us with AIR removed, no longer having cats, and running speed density with an $8D maskon a 730 ECM:

Is it safe to say if we drop the three closed loop O2 constants listed above by 10mv, (maybe 20) that we aren't going to hurt anything, and will help lean the raw fuel stinkiness at idle?


Trying to get a clue, Dave (and may Grumpy rest in peace)

 

Won't hurt anything, it will barely make a change. You probably won't see any difference actually. The 'range' of a narrowband sensor is so small that unless you're at the limit of the sensor's range where actual afr can be wildly far off from where you may think it is you might as well be tuning 1/10th of an afr point. And I think this may be why gm programmed the o2 tables the way they did- to increase the amount of time the motor spends outside the limits of what the o2 sensor can 'see' without getting so far out that it can't control the swing.

 

It can make a difference, Overlap has to be considered.
I run mine about 40-50 mv different at idle on 224/230@112 cam. Cleaned up the rich smell at idle allot. Remember what o2 sensors measure, its not how much fuel is burnt.

 

Read this 3 or more times again...
So the 100 mV is subtracted during normal operation so the switching points are actually closer to the 450 mV sensor point, then the removal of AIR devices advice to remove the 100 mV setting to zero is incorrect.
The correct way to correct for removal of the air pump would be to adjust the three tables by 100 mV (and zero the 100 mV)
I've been studying why my milage is not better and found I have the zero in place at 849D (In $8D)
Since I don't run any emmissions parts my thought was to flatten out the o2 tables and tighten up the control to reduce the "rolling AFR" at part throttle.
Am I going in circles by having this at zero?

 

So JP , did you ever resolve this conundrum of whether to zero out the A.I.R. stuff if it's been removed.

What I'm trying to correct is lots of "Open Loop Stinky" at idle before closed loop kicks in. I seem fairly good(by my Smell-o-meter) in closed loop functions, but need to take out fuel in open loop idle.

I'm suspecting I really should be looking at temperature related fuel supply as well. since it is not officially warmed up yet while in open loop.

 

I thinkI may have confused myself as I reread this section again , as I should have said "Is it safe to say if we drop the three closed loop O2 constants listed above by 10mv, (maybe 20) FROM STOCK"

I made my original question sound like I wanted to drop 10-20 from RBob's settings,and I didn't mean that, I meant from stock. I still working on closed loop "stinky", the smell after entering closed loop.

 

Think I'm going to dredge this up from the deep archives of TGO folklore.

I am getting my VE tables in line to BLMS of 128+/-10. Even with BLMs getting honed in, I have WB02 readings mostly down in the 12.x-13.x range during cruising though. I think the discussion in this thread sums the reason for that up. Car is running rich because the crossover tables have too high mVolts readings. I have now lowered them by .05 mVolts to see what it does
(no driving yet). I have lowered these tables...

Loop Closed Param, Slow o2 R/L vs MAP UPPER THRESHOLD
Loop Closed Param, Slow o2 R/L vs MAP LOWER THRESHOLD
Loop Closed Param, FAST o2 R/L THRESHOLD vs MAP

Any good threads on people honing in these values to line up their BLMs with 14.7 or whatever you want to run?

Thanks,

 

5mV (or 0.05 volts, what I assume you meant) is not going to make much of a difference.

When your WBO2 is reading 12.x-13.x, is your NBO2 actually generating crosscounts? That is quite a ways off, and a little tweaking of the upper/lower thresholds is not going to fix that large of a discrepancy.

Are your NBO2 and WBO2 mounted in the same spot? You can't mount a WBO2 after the cat...

 

.05 volts is 50 mVolts. Maybe I typed that wrong last night. Yep, sure enough, I did. Yah, you wouldn't see that on a gauge lol. I mean 50 MVolts

NB and WB. NB in factory manifold location. WB is at the end of the y-pipe(before Cat if one existed).

NB02 was new this year. I need to check on the cross counts, haven't checked that. OK, just looked at the datalog from yesterday. It appears that it is generating crosscounts. The cross counts go from 0 to 250 in about 15-20 seconds it appears. Thats when TC resets them in my graph to 0.

Thoughts?

 

ack, you are of course right. 0.05V = 50mV.

With your WBO2 in the Y, it's reading the entire engine, whereas the NBO2 is reading only one bank. It is pretty common for the averaged AFR on each bank to be slightly different; the LT1's were somewhat known for this, its called split-BLM (they had an O2 sensor in each manifold and tracked BLM for each bank separately, and the BLM's would be different (split) for each bank). So it's possible that measuring in two different spots (one including both banks, one not) is skewing the numbers some.

With that said, I just checked some of your prior posts and saw that you are running a mild cam in this motor. I don't have any experience tuning EFI with anything other than a stock cam, so I would defer to others that have when it comes to the effect upon NBO2 readings. It's a relatively mild cam though, and in a 383 no less, so I would not have thought that it would affect the NBO2 this much. But it's better to defer to those who have tuned mild cams before.

 

Yep, its a mild cam 383.

I may have misspoke on that low a NB02 reading. I realized now I need to wait for the Integrator to even out. My numbers are in the upper 13.x's to 14s. I took out 50 mVolts in those tables and it took it lean in places...others it was perfect. I think I am on the right track to get it honed in...but need to do more than just a blanket change. Car is running good right now and VE tables are very close. Have a couple issues that need to be addressed...but I think I have the right idea on how to dial it in to the right AFR vs BLMs.

 

Was going to post in another thread about this, but realized I had dredged it up a couple months ago...so I'll dredge it up again.

Working with this parameters as above...

Loop Closed Param, Slow o2 R/L vs MAP UPPER THRESHOLD
Loop Closed Param, Slow o2 R/L vs MAP LOWER THRESHOLD
Loop Closed Param, FAST o2 R/L THRESHOLD vs MAP

Have been working on this, by pulling a little out of the 40 kPa(-50 mV) and 20 kPa (-100 mV) areas of those 3 tables, with just little tweaks I can see the results pretty clear on the WB02. I have the 20 and 40 areas tuned in to the 14.X range pretty nicely.

My problem is 30 kPA. I have been pulling mVolts out and it almost has no effect. I finally got it to lean out form 12.7-12.8:1 to around 13.4:1, but I have pulled out 120 mV out and am going to have to go more to get it inline. It just seems odd. The graph for these coefficients will have a dip in this area trying to get the AFR in line.

What am I missing? Should I be lowering the 02 Filter coefficient as well(it is still set at .602 as original Super AUJP is)? Has anyone else seen this issue?

Thanks,

 

ttt


So in short,

AFR is too lean, raise the O2 numbers
AFR is too rich, lower the o2 numbers.

Correct???
I 'm not sure if I understand the distinction as to when to modify the basic three O2 numbers vs when to attack the Map O2 tables instead.

MAP O2 is to target specific MAP areas vs a general change everywhere which would be the basic lower/threshold/upper O2 numbers? The clues aren't quite adding up for me yet.

Can we shed a bit more light on this?

 

Are we speaking of NB or WB 02 sensor feedback to ECU?

If NB the 02 voltage is contantly moving over median .45 to .50 (whatever it is set at). A/F is constantly changing but held tightly to stoich.

It is not rich nor lean it is near stoich.

When in PE the NB02 will still register a set voltage lets say it is .90 mv. There is now no cycling/swing over median. That would indicate rich PE A/F. Conversely .25 is lean PE A/F.

Some will say you can tune with NB most say it is not accurrate.

There is also a Constant to set for desired A/F. Mine is set to 14.1 for E10.

 

Aren't the three basic O2 values for when the engine is in idle?

And the three MAP tables are for off idle?

RBob.

 

That may be the case. But it struck me, essentially can't I get the same end result by modifying the fuel table? If I'm too rich/lean at a certain map/rpm value, then fuel table gets modified.

So what's the distinction in playing with O2 millivolts to get what seems to be (to me) the same result. Is it that the fuel table will never be "correct" until I tell the ECM what my interpretation of a "correct" value is for Stoich?

I'm getting kind of boggled in the brain. (no beers either)

 

The O2 windows are for closed loop, they are the target. Can move the VE table around all you want, but in closed loop the ECM is going to target the O2 window.

See the Tuning Guide Book sticky, chapter 5, '8746/'7740/'8063 ECM, there is a fueling write up that describes this.

RBob.